Humanized CD30-Targeted Chimeric Antigen Receptor T Cells Exhibit Potent Preclinical Activity Against Hodgkin’s Lymphoma Cells

CD30-directed chimeric antigen receptors (CARs) with single chain antibody fragment (scFv)-binding domains from murine HRS3 show strong cytotoxicity to Hodgkin’s Lymphoma cells and have been used in clinical trials. However, murine scFv in CAR might induce specific rejective immune responses in patients, which compromises the therapeutic effects. The use of human or humanized antibody fragments for CAR construction, rather than those derived from mouse antibodies, can reduce the immunogenicity of the CAR. Importantly, this strategy might simultaneously decrease the risk of cytokine-mediated toxicities and improve CAR T cell persistence. Murine HRS3 antibody has been successfully humanized by grafting the complementarity-determining regions (CDRs) from the mouse antibody framework onto human immunoglobulin consensus sequences, followed by an in vitro evolutionary strategy to select functional Fab fragments with the same affinity as murine sources. In this study, humanized scFvs were utilized to construct a CD30-directed CAR (hHRS3-CAR), and its effectiveness was compared with that of HRS3-CAR. The hHRS3-CAR-T cells specifically kill CD30-positive tumor cell lines in vitro and eliminate lymphoma xenografts in immunodeficient mice with comparable efficiency to HRS3-CAR. The hHRS-CAR-T could be used in clinical trials based on the previously reported advantages of humanized CARs, such as the reduction of immune rejection and better persistence of cells.


INTRODUCTION
Chimeric antigen receptors (CARs) are artificial proteins whose basic structure is composed of an antigen recognition ectodomain and activation endodomain linked by a spacer and transmembrane domain (Eshhar et al., 1993;Sadelain et al., 2013;Boyiadzis et al., 2016). These are engineered receptors used to specifically direct patient T cells to target tumor cells . Adoptive cell therapy with the cluster of differentiation 19 (CD19)-directed chimeric antigen receptor T-cells (CAR-T) has induced durable clinical efficacy and has been approved for clinical first-line treatment, which provides a rationale for the development of other CARs (Kochenderfer et al., 2012;Boyiadzis et al., 2016;Turtle et al., 2016b).
Tumor necrosis factor receptor superfamily member 8 (TNFRSF8) is a 120-kDa type I transmembrane glycoprotein, more commonly referred to as CD30 (van der Weyden et al., 2017). In the normal physiological state, cell surface expression of CD30 is limited to activated T, B, and natural killer lymphocytes (Wasik et al., 2013). However, it is also strongly expressed in malignant hematopoietic cells, including Hodgkin's lymphoma (HL), anaplastic large cell lymphoma (ALCL), primary cutaneous ALCL, lymphomatoid papulosis, and certain cases of transformed mycosis fungoides (Schirrmann et al., 2014;van der Weyden et al., 2017). Thus, it was regarded as an ideal therapeutic target and a wide range of agents were developed for CD30-positive malignancies (Schwarting et al., 1989;Younes et al., 2012). Brentuximab vedotin, a CD30 antibody-drug conjugates, has achieved remarkable results against HL and ALCL (Pro et al., 2012;Younes et al., 2012). Recently, it was shown that anti-CD30 CAR-T cells can induce partial or complete remission in HL and ALCL (Wang et al., 2017;Ramos et al., 2020).
Although CD30 remains an excellent target for CAR-T therapy of HL, various obstacles should be overcome to clinically improve their efficacy, such as the persistence of CAR-T cells, trafficking to tumors, and increased tumor cytotoxicity (Ramos et al., 2017;Hong et al., 2018;Grover and Savoldo, 2019;Hombach et al., 2019). The murine anti-CD30 monoclonal antibody (HRS3), and HRS3-based bispecific antibody (AFM13) and recombinant fusion protein combined with cytokines, and CAR-T have been developed and applied to the specific immunotherapy of Hodgkin's lymphoma, including CAR-T therapy (Hombach et al., 1993;Renner et al., 2000;Jahn et al., 2012;Grover and Savoldo, 2019). Currently, most single chain antibody fragments (scFvs) are murine-derived, which, owing to immunogenicity, may weaken the efficacy of the treatment (Zou et al., 2018). One strategy to reduce the immunogenicity of CAR-T cells is to humanize scFv. Humanized CAR-T cells have also received attention because they are less antigenic and have a long survival time in vivo (Turtle et al., 2016a;Sommermeyer et al., 2017;Rafiq et al., 2020). Starting from the previously characterized cognate HRS3 mouse monoclonal antibody, the bacterially produced functional Fab fragment was humanized by grafting the CDRs from the mouse antibody framework onto human immunoglobulin consensus sequences, followed by an in vitro evolutionary strategy to select functional Fab fragments with the same affinity as murine sources (Schlapschy et al., 2004). The humanized HRS3 antibody is fully functional in recognizing its native receptor antigen. CAR-T based on humanized HRS3-scFv has not been reported.
In this study, murine and humanized HRS3 scFv-based CARs (HRS3-CAR and hHRS3-CAR, respectively) were constructed. We used humanized scFvs to construct hHRS3-CARs, and its effectiveness was compared with that of HRS3-CAR.

Cell Culture
L428 and L540 Hodgkin's lymphoma-derived cell lines were purchased from Shanghai Fudan-Zhangjiang Bio-Pharmaceutical Co., Ltd. (Shanghai, China). Cells were cultured in RPMI-1640 (Hyclone, United States) supplemented with 20% fetal bovine serum (FBS; Gibco). Raji cells were purchased from ATCC and cultured in RPMI-1640 medium containing 10% FBS. HEK293T cells were cultured in Dulbecco's modified Eagle's medium (Millipore, United States) supplemented with 10% FBS at 37°C and 5% CO 2 . The cells were authenticated by via Short Tandem Repeat (STR) profiling and tested negative for mycoplasma contamination.

Construction of Plasmids Encoding anti-CD30 CARs
The CD19-specific single-chain antibody CD19-scFv was derived from anti-CD19 mAb as previously described (Yang et al., 2017). The VL and VH regions of murine and humanized HRS3 and HRS3 antibodies were obtained from the published article (Schlapschy et al., 2004). The extracellular domains containing VL and VH regions were synthesized and cloned into a lentiviral backbone containing other parts of CAR: a CD8 hinge spacer and transmembrane region, 4-1BB, and CD3ζ endo-domains under the control of the CMV promoter. The resulting plasmids were named HRS3-CAR and hHRS3-CAR, respectively.

Lentiviral Preparation and Transduction of T-Cells
Lentiviral DNA vectors were transfected with Lipo6000 ™ Transfection Reagent (Beyotime, China) according to the manufacturer's protocol. For lentivirus production, 20 μg of core plasmid together with helper plasmids (10 μg pCMVΔ8.9 and 4 μg pMD2. G) were transfected into 293T cells as described previously . Viral supernatants were collected 48 and 72 h after transfection and filtered through a 0.45 μm filter. After centrifugation at 25,000 rpm for 2.5 h at 4°C, the virus was suspended in 0.1% bovine serum albumin (BSA) in PBS, dissolved overnight and stored in aliquots at 80°C.
Primary human T cells were isolated from healthy human blood, as previously described . T cells were cultured in advanced RPMI 1640 medium (Life Technologies, United States) containing 10% FBS (Life Technologies, United States) with 200 U/ml IL-2 (PeproTech, United States). T cells were activated by adding Dynabeads human T-activator CD3/CD28 kit (Life Technologies, United States) according to the manufacturer's instructions. After 48 h, at a 1:1 ratio, and on day 3, T cells were transduced with lentivirus (Multiplicity of infection, MOI 20) in the presence of Lentiboost (Sirion Biotech), followed by a medium change after 24 h transfer to 24-well plates.

Cytotoxicity Assay
Cytotoxicity assays were conducted using a slightly modified version of a previously described assay, and cytokine concentrations were determined using enzyme-linked immunosorbent assay (ELISA) . For data analysis, target cells were labeled with carboxyfluorescein succinimidyl ester (CFSE) and dead target cells were identified as double-positive for CFSE and propidium iodide (PI) using a FACS Canto flow cytometer. For each sample, 10000 events of P1 were collected. The percentage of specific target cell death was calculated as follows: 100 ×

Xenograft Mouse Model
The protocol was approved by the Ethics Committee of the West China Hospital of Sichuan University West China Hospital.
Six-to eight-week-old female NOD-Prkdcem26ll2rgem26Nju (NCG) mice were purchased from GemPharmatech Co., Ltd (Jiangsu, China) and engrafted with 2 × 10 5 L428-EGFP-luci cells via tail vein injection. Five days later, CAR-or control T cells that had been expanded in vitro for 7 days were injected intravenously. The mice were serially imaged using 2D bioluminescence imaging as previously described to determine tumor progression (Guercio et al., 2021). All captured images were analyzed using Living Image software 4.1 (PerkinElmer, United States). The animals were examined every day for survival and sacrificed when moribund for collection of tissue samples. At the end of the experiment (day 64), the remaining surviving mice were euthanized and animal samples were collected. Each experiment included 6 mice per group and was repeated twice (total n 12 mice per group).

Cytokine Secretion Assays
Effector cells (2.5 × 10 5 NC, CD19-CAR, HRS3-CAR, or hHRS-CAR-T cells) and target cells (5 × 10 4 ) were co-cultured in the absence of IL-2 at a ratio of 5:1 in 96-well plates for 24 h. Supernatants were collected and subjected to ELISA assay (BD Biosciences, United States) measurements according to the manufacturer's instructions.

Statistical Analysis
Graphs were plotted using the GraphPad Prism 8.0 statistical software. Data were analyzed using SPSS 22.0 software (IBM, United States). Statistical differences between two groups were analyzed using the unpaired Student's t-test with Welch correction. Unless otherwise stated, data are presented as mean ± SD, and statistical significance was set at p < 0.05.

Humanized HRS3-CAR-T Cells Exert a Similar In vitro Anti-Lymphoma Activity Compared to HRS3-BBz T-Cells
We constructed hHRS3-CAR and HRS3-CAR for an initial validation of the humanized CAR, by cloning the scFvs of hHRS3 or HRS3, respectively, into the same CAR scaffold ( Figure 1A) (Schlapschy et al., 2004). The CAR expressing anti-CD30(HRS3)-scFv was used as the positive control and that expressing anti-CD19-scFv was used as the negative control (CD19-CAR), as previously described . Comparing the amino acid sequences of VL and VH for HRS3-scFv and hHRS3-scFv, it was found that there were no significant changes in the amino acids in the VL; however, amino acid changes in the VH were apparent ( Figure 1B).
To verify the function of hHRS-CAR-T, we selected non-HL cells Raji and U937, and the HL cells L428 and L540. We evaluated CD30 protein expression using flow cytometry. CD30 was not detected in Raji or U937 cells, but was detected in the HL cell lines (Figure 2A, Supplementary Figure S1A). Next, we examined CAR expression and evaluated the ability of CAR-T cells to mediate the lysis of CD30 + HL cells ( Figure 2B). CAR-T cells lysed CD30 + L428 and L540 cells ( Figure 2C), but not Raji and U937 cells (Supplementary Figure S1B). We assessed the cytotoxicity of CD30-CAR T cells using 24-h cytotoxicity assays ( Figure 2D). T cells expressing CD19-CAR served as negative controls. CAR-T cells specifically recognized and eliminated HL cells in an effector-to-target ratio-dependent manner ( Figures 2E,F).

Humanized HRS3-CAR-T Specifically Kills CD30-Positive Cells Similar to HRS3-CAR-T
To test whether T cells expressing hHRS3-CAR were capable of specifically recognizing tumor lines expressing CD30, we coincubated CAR-T cells with Raji cells expressing CD30 and those that did not ( Figure 2G). CAR-T cells specifically lysed Raji-CD30 cells, but not Raji cells ( Figure 2H). The amounts of secreted cytokines, IFN-γ and IL-2 ( Figure 2I, Supplementary Figure S2), were measured. The hHRS3-CAR-T cells recognized all CD30 + tumor lines and secreted high levels of IFN-γ. Very low levels of IFN-γ were observed when hHRS3-CAR-T cells were cocultured with Raji cells. In contrast, higher levels of IFN-γ were observed when CD19-CAR-T cells were co-cultured with Raji cells.

Expanded hHRS3-CAR-T Cells Express a Higher Central Memory Phenotype
To identify the effect of CAR expression on T cell phenotypes, markers of the T cell subsets were analyzed using FACS. The CD4 to CD8 ratio of CAR + T cells was not significantly different between the HRS3-CAR-T and hHRS3-CAR-T cells ( Figures  3A,B). Notably, participating effector memory (Tem) cells were significantly enriched in CD30-CAR-T cells 4 days after infection, whereas central memory (Tcm) cells were significantly enriched in CAR-T cells on the eighth day after infection (Figures 3C,D). The hHRS3-CAR-T cells maintained a higher proportion of CCR7 + CD45RO + Tcm cells, while HRS3-CAR-T cells had larger populations of CCR7 − CD45RO + Tem cells compared to their counterparts. Humanization of the HRS3 VL domain required 18 amino acid exchanges while retaining one amino acid from the murine framework sequence, whereas for the HRS3 VH domain, 27 amino acids were exchanged and four murine residues were retained (Schlapschy et al., 2004). The figure shows the schematic diagram of amino acids after mutation.

Antitumor Responses by Humanized HRS3 Chimeric Antigen Receptor T Cells in vivo
To evaluate the anti-tumor activity of hHRS3-CAR-T and HRS3-CAR-T cells in murine models, NCG mice (six per group) were injected intravenously with L428 tumor cells (2 × 10 6 cells per mouse). After 5 days (day 0), the tumor-bearing mice were imaged using a bioluminescence imaging system for Luc expression, and the mice were randomly divided into four groups for different treatments ( Figures 4A,B). NC T cells, CD19-CAR-T cells, HRS3-CAR-T cells, and hHRS3-CAR-T cells (1 × 10 7 cells per mouse) were infused intravenously as a single dose. hHRS3-CAR-T cells showed excellent efficacy comparable to HRS3-CAR cells. In contrast, CD19-CAR-T and NC T cells exhibited progressive tumor growth ( Figure 4C). When tumor sizes were assessed 21 days after CAR-T cell infusion, we found that T cells expressing hHRS3-CAR were more effective at reducing tumor sizes than T cells expressing HRS3-CAR ( Figure 4D); however, survival and weight change were not statistically different when treatment with hHRS3-CAR-T and HRS3-CAR-T cells was compared. Mice in the control group died after 40 days, whereas mice in the HRS3-CAR-T and hHRS3-CAR-T groups died after 63 days. Similar relative survival curves were obtained in repeated experiments, and the survival curve data from both experiments were combined for analysis ( Figure 5A). Owing to significant liver ascites and death in the control group, body weight changes were only recorded for 35 days (Supplementary Figure S3). Previous studies have reported a class of HL tumor models that preferentially caused accumulation of tumors in the liver abdomen (Guercio et al., 2021). Macroscopic analysis of the sacrificed mice showed large tumor masses located preferentially in the liver ( Figure 5B). Therefore, the anti-tumor activity of hHRS3-CAR-T cells was comparable to that of HRS3-CAR-T cells in this model. Considering that hHRS3-CAR Tcm cells were significantly enriched in vitro, we evaluated its proportion in vivo. The results showed that the difference in Tcm cell ratio was not statistically significant in the blood, spleen, and bone marrow of mice treated with hHRS3-CAR-T cells versus HRS3-CAR ( Figures 5C-E).

DISCUSSION
To date, several clinical trials have been conducted in adults with relapsed/refractory CD30 + lymphoma using different CD30-CAR-T cells (Di Stasi et al., 2009;Wang et al., 2017). Compared with the great success achieved by CD19-redirected CAR, CD30-target CAR still has room for improvement. Previous studies have demonstrated that the transfer of autologous human T-cells expressing foreign proteins, including CARs derived from murine scFvs, can elicit cellular and humoral immune responses that may compromise CAR-T effects (Lamers et al., 2006;Wang et al., 2020;Wagner et al., 2021). The use of human or humanized antibody fragments for CAR construction, rather than those derived from mouse antibodies, such as the production of modified CD19 and Her-2 CAR-T, combination of humanized or fully human fragments, and modification of the extracellular hinge region and/or transmembrane domain, can reduce the immunogenicity of the CAR (Hombach et al., 2010;Jonnalagadda et al., 2015). Importantly, this strategy might simultaneously decrease the risk of cytokine-mediated toxicity and improve CAR-T cell persistence (Sommermeyer et al., 2017;Rafiq et al., 2020). HRS3 is derived from a murine antibody (da Costa et al., 1992;Hombach et al., 1998); humanized HRS3 scFv was derived from HRS3 mAb, which may represent another class of anti-CD30 antibodies (Schlapschy et al., 2004). Notably, previous studies have demonstrated that the humanized Fab fragment was fully functional with respect to CD30 binding; however, whether the CAR possessing this molecule as an extracellular recognition domain can safely and effectively eliminate target cells remains requires further investigation. In this study, we found that hHRS3-CAR-T cells were highly efficacious against antigen-specific tumor cells in both in vitro and in vivo with responses that were comparable to those of HRS3-CAR-T cells. Besides scFv, other elements of CAR may also produce rejection. In this study, the sequence of the hinge spacer and transmembrane domain in the CAR were all derived from human CD8α to minimize immunogenicity. Furthermore, the linker of (GGGGS)3 that connects the heavy and light chains in the scFv has been widely used in antibodies and CARs. However, these strategies do not completely eliminate immunogenicity. Even a fully human antibody fragment might also lead to immune reactions in the human; therefore, to conclusively determine whether they are nonimmunogenic requires their application in human clinical trials.
Retrospective analysis of published CAR-T cell clinical studies revealed that an elevated proportion of Tcm or less differentiated CAR-T cells provided superior antitumor efficacy, and Tcm-derived CAR-T cells are functionally superior to those made from bulk CD8 + T cells (Chang et al., 2015;Tao et al., 2020). Tcm provides effective long-term memory responses because they have the capacity to persist long term in the circulation, have a high proliferative capacity, and can replenish other memory T-cell subsets, including Tem (Gehad et al., 2018;Mauriello et al., 2020). Our results showed that hHRS3-CAR could produce the higher proportion of Tcm compared to parental CAR. Similar results were also observed in previous reports (Zhao et al., 2019). However, the mechanism underlying this effect remain to be elucidated up to now. We speculate that it is due to the continuous stimulation caused by antigen, such as murine scFv, accelerating the differentiation of Tcm to Tem. Although the advantages of humanized CAR-T have been proven in clinical trials, the improvement was not significant in our experiment, which may be due to the huge differences in the immune environment in immunodeficient mice and human. For example, the humanized anti-CD19 CAR did not show a difference in the ability to kill target cells in vitro and in mice compared to parental CAR, but it exhibited significantly lower side effects in the patients with B-cell lymphoma (Heard et al., 2021).
In conclusion, we focused on developing a CD30-specific CAR using humanized HRS3-scFv. hHRS3-CAR-T cells specifically killed CD30 + tumor cells in vitro with efficiency similar to that of murine HRS3-CAR-T, and these T cells significantly inhibited the established CD30 + human lymphoma tumor cells in immunodeficient mice. These findings indicate that hHRS3-CAR-T cells are a promising cellular therapeutic agent and warrants further clinical exploration.
FIGURE 5 | Central memory cells (Tcm) expressing hHRS3-CAR were enriched in mice and conferred long-lasting immunity to HL. (A) Survival of mice is shown as a Kaplan-Meier curve (data shown combined from two independent experiments, n 12/group). Survival was not statistically different for NC mice versus CD19; p 0.8037. p < 0.0001 for the comparison of HRS3 group versus CD19 mice. p < 0.0001 for the comparison of hHRS3 group versus CD19 mice. NC, non-transduced T cells. (B) Flow cytometry detection of tumor cell involvement in the liver of mice in each group. The control group was CD19 group (n 3), and the experimental group was HRS3 and hHRS3 groups (n 4). (C) Spleen, (D) blood, and (E) bone marrow from surviving CD30-CAR-treated mice (n 3) were analysed for the number of CAR-T cell subsets using flow cytometry (mean ± SD). T cells are functionally divided into four subsets: naïve (CCR7 + CD45RO − ); central memory (CCR7 + CD45RO + ); effector memory (CCR7 − CD45RO + ); and terminal effector (CCR7 − CD45RO − ), according to the cell surface expression of CCR7 and CD45RO. Each experiment included 6 mice per group and was repeated twice (total n 12 mice per group). (B) data shown are representative or (C-E) combined from two independent experiments (n 12). * p-value < 0.05, ** p-value < 0.01, *** p-value < 0.001 vs control (CD19).
CAR-T Based on Humanized HRS3-scFv

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding authors.

ETHICS STATEMENT
The animal study was reviewed and approved by the Animal Ethics Committee of West China Hospital of Sichuan University.

AUTHOR CONTRIBUTIONS
JG, DY, and XZ: conception and design. JG and DY: methodology, validation, data collection, analysis, writingoriginal draft and editing. SH and WY: methodology, validation, data collection, analysis. YZ and XZ: administrative, writing, or material support.

FUNDING
This work was supported by 1.3.5 project for disciplines of excellence, West China Hospital, Sichuan University (ZYYC20002) and the Department of Science and Technology of Sichuan Province (2020JDRC0019).